DSBICNet: dynamic-static bidirectional interaction collaborative network for micro-expression recognition

Key Points

• Micro-expression recognition accuracy improved significantly through introduced dynamic-static bidirectional interaction.
• An overall accuracy of 92% was achieved on the tested dataset of micro-expressions.
• The analysis used a deep learning approach to extract relevant features for effective recognition.
• The study suggests that innovative frameworks may enhance recognition tasks in computer vision applications.